DOCSLIB.ORG

Arxiv:1709.02341V5 [Q-Bio.NC] 11 Oct 2018

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Arxiv:1709.02341V5 [Q-Bio.NC] 11 Oct 2018 DEEP ACTIVE INFERENCE KAI UELTZHOFFER¨ Abstract. This work combines the free energy principle from cognitive neuroscience and the ensuing active inference dynamics with recent advances in variational inference on deep generative models and evolution strategies as efficient large-scale black-box optimisation technique, to introduce the "deep active inference" agent. This agent tries to minimize a variational free energy bound on the average surprise of its sensations, which is motivated by a homeostatic argument. It does so by changing the parameters of its generative model, together with a variational density approximating the posterior distribution over latent variables, given its observations, and by acting on its environment to actively sample input that is likely under its generative model. The internal dynamics of the agent are implemented using deep neural networks, as used in machine learning, and recurrent dynamics, making the deep active inference agent a scalable and very flexible class of active inference agents. Using the mountaincar problem, we show how goal-directed behaviour can be implemented by defining sensible prior expectations on the latent states in the agent's model, that it will try to fulfil. Furthermore, we show that the deep active inference agent can learn a generative model of the environment, which can be sampled from to understand the agent's beliefs about the environment and its interaction with it. 1. Introduction Active Inference (Friston et al., 2006, 2010; Friston, 2012) is a normative theory of brain function derived from the properties required by active agents to survive in dynamic, fluctuating environments. This theory is able to account for many aspects of action and perception as well as anatomic and physiologic features of the brain on one hand (Brown and Friston, 2012; Friston, 2005; Friston et al., 2011; Friston and Kiebel, 2009; Schwartenbeck et al., 2015; Adams et al., 2013), and encompasses many formal theories about brain function on the other (Friston, 2010). In terms of its functional form it rests on the minimisation of an upper variational bound on the agents average surprise. In this way it is formally very similar to state of the art algorithms for variational inference in deep generative models (Rezende et al., 2014; Kingma and Welling, 2014; Chung et al., 2015). However, optimising this bound for active agents introduces a dependency on the true dynamics of the world, to which the agent usually does not have access, and whose true arXiv:1709.02341v5 [q-bio.NC] 11 Oct 2018 Date: October 12, 2018 This is a pre-print of an article published in Biological Cybernetics. The final authenticated ver- sion is available online at: https://doi.org/10.1007/s00422-018-0785-7 The author has archived a personal copy of the accepted manuscript at his personal homepage: https://kaiu.me. 1 2 KAI UELTZHOFFER¨ functional form does not have to coincide with the functional form of the agent's generative model. Here we solve these problems using deep neural networks (LeCun et al., 2015) and recurrent neural networks (Karpathy et al., 2015) as flexible function approximators, which allow the agent's generative model to learn a good approximation of the true world dynamics. Futhermore we apply evolution strategies (Salimans et al., 2017) to estimate gradients on the variational bound, averaged over a population of agents. This formalism allows to obtain gradient estimates even for non-differentiable objective functions, which is the case here; since the agent does neither know the equations of motions of the world, nor its partial derivatives. In this way, our approach pairs active inference with state of the art machine learning techniques to create agents that can successfully reach goals in complex environments while simultaneously building a generative model of themselves and their surroundings. In this work we want to lay out the basic form of these so called "Deep Active Inference" agents and illustrate their dynamics using a simple, well known problem from reinforcement learning, namely the mountain car problem (Moore, 1991). By utilising models and optimisation techniques that have been applied successfully to real-world data and large-scale problems (Rezende et al., 2014; Kingma and Welling, 2014; Chung et al., 2015; Salimans et al., 2017), our agent can be scaled to complex and rich environments. With this paper we publish the full implementation of the resulting Deep Active Inference agent, together with all scripts used to create the figures in this publication at https://www.github.com/ kaiu85/deepAI_paper. In section 2 we will briefly recapitulate the active inference principle. In section 3 we will describe the mountain car environment and introduce a deep active inference agent that is able to solve this problem. In section 4 we will show that the agent can solve the mountain car problem while simultaneously learning a generative model of its environment. In section 5 we will discuss possible further directions of this approach and its relation to other approaches. 2. Active Inference Active Inference (Friston et al., 2006, 2010; Friston, 2012) rests on the basic assumption that any agent in exchange with a dynamic, fluctuating environment has to keep certain inner parameters within a well defined range. Otherwise, it would sooner or later encounter a phase transition due to which it would loose its defining characteristics and therefore disappear. Thus, the agent must restrict itself to a small volume in its state space. This can be formalised using the entropy of the probability distribution p(s∗) of finding the agent in a given state s∗ of its state space S∗: Z H(S∗) = ( ln p(s∗)) p(s∗) ds∗ s∗2S∗ − By minimising this entropy, an agent can counter dispersive effects from its environment and maintain a stable identity in a fluctuating environment. However, an agent does not have direct access to an objective measurement of its current state. Instead it only perceives itself and the world around it via its sensory epthelia. This DEEP ACTIVE INFERENCE 3 can be described by a potentially noisy sensory mapping o = g(s) from states s to sensations o. Defining the sensory entropy Z H(O) = ( ln p(o)) p(o) do o2O − over the space O of all possible observations of an agent, one can derive the following inequality in the absence of sensory noise (Friston et al., 2010) Z ∗ ∗ H(O) H(S ) + p(s ) ln gs∗ ds ≥ s∗2S∗ j j Agents, whose sensory organs did not have a good mapping of relevant physical states to appropriate sensory inputs, would not last very long. So the mapping between the agents true states and its sensations is assumed to have an almost constant sensitivity, in terms of the determinant of the Jacobian gs∗ over the encountered range of states s. This makes the last term approximately constantj j and allows upper bounding the entropy of the agent's distribution on state space by the entropy of its sensory states plus this constant term (Friston et al., 2010). Thus, to ensure keep its physiological variables within well-defined bounds, an agent has to minimize its sensory entropy H(O) 1 Assuming ergodicity, i.e. the equivalence of time- and ensemble-averages, one can write the sensory entropy as 1 Z T H(O) = lim ln p(o(t)) dt T !1 −T 0 From the calculus of variations it follows that an agent can minimize its sensory entropy by minimising its sensory surprise ln p(o(t)) at all times, in terms of the following Euler- Lagrange-equation: − o ( ln p(o(t))) = 0 r − To be able to efficiently do this, our agent needs a statistical model of its sensory inputs, to evaluate p(o). Since the world in which we live is hierarchically organised, dynamic, and features a lot of complex noise, we assume that the agent's model is a deep, recur- rent, latent variable model (Conant and Ashby, 1970). Furthermore we assume that this model is generative, using the observation that we are able to imagine certain situations and perceptions (like the image of a blue sky over a desert landscape) without actually experiencing or having experienced them. Thus, we work with a generative model pθ(o; s) of sensory observations o and latent variables s, that represent the hidden true states of the world, which we can factorise into a likelihood function pθ(o s) and a prior on the states j pθ(s): pθ(o; s) = pθ(o s)pθ(s) j The set θ comprises the slowly changing parameters that the agent can change to improve its model of the world. In the brain this might be the pattern and strength of synapses. 1In more general formulations of active inference, the assumption that the mapping between hidden states and outcomes is constant can be relaxed (Friston et al., 2015). 4 KAI UELTZHOFFER¨ Given this factorisation, to minimize surprise, the agent has to solve the hard task of calculating Z pθ(o) = pθ(o s)pθ(s) ds j by marginalising over all possible states that could lead to a given observation. As the dimensionality of the latent state space S can be very high, this integral is extremely hard to solve. Therefore a further assumption of the free energy principle is, that agents do not try to minimize the surprise ln pθ(o) directly, but rather minimize an upper bound, which is a lot simpler to calculate.− Using the fact that the Kullback-Leibler (KL) Divergence Z pa(x) DKL(pa(x) pb(x)) = ln pa(x) dx jj x2X pb(x) between two arbitrary distributions pa(x) and pb(x) with a shared support on a com- mon space X is always greater or equal to zero, and equal to zero if and only if the two distributions are equal, we can define the variational free energy as: F (o; θ; u) = ln pθ(o) + DKL(qu(s) p(s o)) ln pθ(o) − jj j ≥ − Here qu(s) is an arbitrary, so called variational density over the space of hidden states s, which belongs to a family of distributions parameterised by a time-dependent, i.e.
Load more

Recommended publications
  • Deep Active Inference for Autonomous Robot Navigation
    Published as a workshop paper at “Bridging AI and Cognitive Science” (ICLR 2020) DEEP ACTIVE INFERENCE FOR AUTONOMOUS ROBOT NAVIGATION Ozan C¸atal, Samuel Wauthier, Tim Verbelen, Cedric De Boom, & Bart Dhoedt IDLab, Department of Information Technology Ghent University –imec Ghent, Belgium [email protected] ABSTRACT Active inference is a theory that underpins the way biological agent’s perceive and act in the real world. At its core, active inference is based on the principle that the brain is an approximate Bayesian inference engine, building an internal gen- erative model to drive agents towards minimal surprise. Although this theory has shown interesting results with grounding in cognitive neuroscience, its application remains limited to simulations with small, predefined sensor and state spaces. In this paper, we leverage recent advances in deep learning to build more complex generative models that can work without a predefined states space. State repre- sentations are learned end-to-end from real-world, high-dimensional sensory data such as camera frames. We also show that these generative models can be used to engage in active inference. To the best of our knowledge this is the first application of deep active inference for a real-world robot navigation task. 1 INTRODUCTION Active inference and the free energy principle underpins the way our brain – and natural agents in general – work. The core idea is that the brain entertains a (generative) model of the world which allows it to learn cause and effect and to predict future sensory observations. It does so by constantly minimising its prediction error or “surprise”, either by updating the generative model, or by inferring actions that will lead to less surprising states.
    [Show full text]
  • Allostasis, Interoception, and the Free Energy Principle
    1 Allostasis, interoception, and the free energy principle: 2 Feeling our way forward 3 Andrew W. Corcoran & Jakob Hohwy 4 Cognition & Philosophy Laboratory 5 Monash University 6 Melbourne, Australia 7 8 This material has been accepted for publication in the forthcoming volume entitled The 9 interoceptive mind: From homeostasis to awareness, edited by Manos Tsakiris and Helena De 10 Preester. It has been reprinted by permission of Oxford University Press: 11 https://global.oup.com/academic/product/the-interoceptive-mind-9780198811930?q= 12 interoception&lang=en&cc=gb. For permission to reuse this material, please visit 13 http://www.oup.co.uk/academic/rights/permissions. 14 Abstract 15 Interoceptive processing is commonly understood in terms of the monitoring and representation of the body’s 16 current physiological (i.e. homeostatic) status, with aversive sensory experiences encoding some impending 17 threat to tissue viability. However, claims that homeostasis fails to fully account for the sophisticated regulatory 18 dynamics observed in complex organisms have led some theorists to incorporate predictive (i.e. allostatic) 19 regulatory mechanisms within broader accounts of interoceptive processing. Critically, these frameworks invoke 20 diverse – and potentially mutually inconsistent – interpretations of the role allostasis plays in the scheme of 21 biological regulation. This chapter argues in favour of a moderate, reconciliatory position in which homeostasis 22 and allostasis are conceived as equally vital (but functionally distinct) modes of physiological control. It 23 explores the implications of this interpretation for free energy-based accounts of interoceptive inference, 24 advocating a similarly complementary (and hierarchical) view of homeostatic and allostatic processing.
    [Show full text]
  • Neural Dynamics Under Active Inference: Plausibility and Efficiency
    entropy Article Neural Dynamics under Active Inference: Plausibility and Efficiency of Information Processing Lancelot Da Costa 1,2,* , Thomas Parr 2 , Biswa Sengupta 2,3,4 and Karl Friston 2 1 Department of Mathematics, Imperial College London, London SW7 2AZ, UK 2 Wellcome Centre for Human Neuroimaging, Institute of Neurology, University College London, London WC1N 3BG, UK; [email protected] (T.P.); [email protected] (B.S.); [email protected] (K.F.) 3 Core Machine Learning Group, Zebra AI, London WC2H 8TJ, UK 4 Department of Bioengineering, Imperial College London, London SW7 2AZ, UK * Correspondence: [email protected] Abstract: Active inference is a normative framework for explaining behaviour under the free energy principle—a theory of self-organisation originating in neuroscience. It specifies neuronal dynamics for state-estimation in terms of a descent on (variational) free energy—a measure of the fit between an internal (generative) model and sensory observations. The free energy gradient is a prediction error—plausibly encoded in the average membrane potentials of neuronal populations. Conversely, the expected probability of a state can be expressed in terms of neuronal firing rates. We show that this is consistent with current models of neuronal dynamics and establish face validity by synthesising plausible electrophysiological responses. We then show that these neuronal dynamics approximate natural gradient descent, a well-known optimisation algorithm from information geometry that follows the steepest descent of the objective in information space. We compare the information length of belief updating in both schemes, a measure of the distance travelled in information space that has a Citation: Da Costa, L.; Parr, T.; Sengupta, B.; Friston, K.
    [Show full text]
  • 15 | Variational Inference
    15 j Variational inference 15.1 Foundations Variational inference is a statistical inference framework for probabilistic models that comprise unobserv- able random variables. Its general starting point is a joint probability density function over observable random variables y and unobservable random variables #, p(y; #) = p(#)p(yj#); (15.1) where p(#) is usually referred to as the prior density and p(yj#) as the likelihood. Given an observed value of y, the first aim of variational inference is to determine the conditional density of # given the observed value of y, referred to as the posterior density. The second aim of variational inference is to evaluate the marginal density of the observed data, or, equivalently, its logarithm Z ln p(y) = ln p(y; #) d#: (15.2) Eq. (15.2) is commonly referred to as the log marginal likelihood or log model evidence. The log model evidence allows for comparing different models in their plausibility to explain observed data. In the variational inference framework it is not the log model evidence itself which is evaluated, but rather a lower bound approximation to it. This is due to the fact that if a model comprises many unobservable variables # the integration of the right-hand side of (15.2) can become analytically burdensome or even intractable. To nevertheless achieve its two aims, variational inference in effect replaces an integration problem with an optimization problem. To this end, variational inference exploits a set of information theoretic quantities as introduced in Chapter 11 and below. Specifically, the following log model evidence decomposition forms the core of the variational inference approach (Figure 15.1): ln p(y) = F (q(#)) + KL(q(#)kp(#jy)): (15.3) In eq.
    [Show full text]
  • Respiration, Neural Oscillations, and the Free Energy Principle
    fnins-15-647579 June 23, 2021 Time: 17:40 # 1 REVIEW published: 29 June 2021 doi: 10.3389/fnins.2021.647579 Keeping the Breath in Mind: Respiration, Neural Oscillations, and the Free Energy Principle Asena Boyadzhieva1*† and Ezgi Kayhan2,3† 1 Department of Philosophy, University of Vienna, Vienna, Austria, 2 Department of Developmental Psychology, University of Potsdam, Potsdam, Germany, 3 Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany Scientific interest in the brain and body interactions has been surging in recent years. One fundamental yet underexplored aspect of brain and body interactions is the link between the respiratory and the nervous systems. In this article, we give an overview of the emerging literature on how respiration modulates neural, cognitive and emotional processes. Moreover, we present a perspective linking respiration to the free-energy principle. We frame volitional modulation of the breath as an active inference mechanism Edited by: in which sensory evidence is recontextualized to alter interoceptive models. We further Yoko Nagai, propose that respiration-entrained gamma oscillations may reflect the propagation of Brighton and Sussex Medical School, United Kingdom prediction errors from the sensory level up to cortical regions in order to alter higher level Reviewed by: predictions. Accordingly, controlled breathing emerges as an easily accessible tool for Rishi R. Dhingra, emotional, cognitive, and physiological regulation. University of Melbourne, Australia Georgios D. Mitsis, Keywords: interoception, respiration-entrained neural oscillations, controlled breathing, free-energy principle, McGill University, Canada self-regulation *Correspondence: Asena Boyadzhieva [email protected] INTRODUCTION † These authors have contributed “The “I think” which Kant said must be able to accompany all my objects, is the “I breathe” which equally to this work actually does accompany them.
    [Show full text]
  • Active Inference Body Perception and Action for Humanoid Robots
    Active inference body perception and action for humanoid robots Guillermo Oliver1, Pablo Lanillos1;2, Gordon Cheng1 Abstract—Providing artificial agents with the same computa- Body perception tional models of biological systems is a way to understand how intelligent behaviours may emerge. We present an active inference body perception and action model working for the first time in a humanoid robot. The model relies on the free energy principle proposed for the brain, where both perception and action goal is to minimise the prediction error through gradient descent on the variational free energy bound. The body state (latent variable) is inferred by minimising the difference between the observed (visual and proprioceptive) sensor values and the predicted ones. Simultaneously, the action makes sensory data sampling to better Action correspond to the prediction made by the inner model. We formalised and implemented the algorithm on the iCub robot and tested in 2D and 3D visual spaces for online adaptation to visual changes, sensory noise and discrepancies between the Fig. 1. Body perception and action via active inference. The robot infers model and the real robot. We also compared our approach with its body (e.g., joint angles) by minimising the prediction error: discrepancy classical inverse kinematics in a reaching task, analysing the between the sensors (visual sv and joint sp) and their expected values (gv(µ) suitability of such a neuroscience-inspired approach for real- and gp(µ)). In the presence of error it changes the perception of its body µ world interaction. The algorithm gave the robot adaptive body and generates an action a to reduce this discrepancy.
    [Show full text]
  • A Free Energy Principle for Biological Systems
    Entropy 2012, 14, 2100-2121; doi:10.3390/e14112100 OPEN ACCESS entropy ISSN 1099-4300 www.mdpi.com/journal/entropy Article A Free Energy Principle for Biological Systems Friston Karl The Wellcome Trust Centre for Neuroimaging, Institute of Neurology, Queen Square, London, WC1N 3BG, UK; E-Mail: [email protected]; Tel.: +44 (0)203 448 4347; Fax: +44 (0)207 813 1445 Received: 17 August 2012; in revised form: 1 October 2012 / Accepted: 25 October 2012 / Published: 31 October 2012 Abstract: This paper describes a free energy principle that tries to explain the ability of biological systems to resist a natural tendency to disorder. It appeals to circular causality of the sort found in synergetic formulations of self-organization (e.g., the slaving principle) and models of coupled dynamical systems, using nonlinear Fokker Planck equations. Here, circular causality is induced by separating the states of a random dynamical system into external and internal states, where external states are subject to random fluctuations and internal states are not. This reduces the problem to finding some (deterministic) dynamics of the internal states that ensure the system visits a limited number of external states; in other words, the measure of its (random) attracting set, or the Shannon entropy of the external states is small. We motivate a solution using a principle of least action based on variational free energy (from statistical physics) and establish the conditions under which it is formally equivalent to the information bottleneck method. This approach has proved useful in understanding the functional architecture of the brain. The generality of variational free energy minimisation and corresponding information theoretic formulations may speak to interesting applications beyond the neurosciences; e.g., in molecular or evolutionary biology.
    [Show full text]
  • A Free Energy Principle for a Particular Physics
    A FREE ENERGY PRINCIPLE FOR A PARTICULAR PHYSICS Karl Friston The Wellcome Centre for Human Neuroimaging, UCL Queen Square Institute of Neurology, London, UK WC1N 3AR. Email: [email protected] (This work is under consideration for publication by The MIT Press) Abstract This monograph attempts a theory of every ‘thing’ that can be distinguished from other ‘things’ in a statistical sense. The ensuing statistical independencies, mediated by Markov blankets, speak to a recursive composition of ensembles (of things) at increasingly higher spatiotemporal scales. This decomposition provides a description of small things; e.g., quantum mechanics – via the Schrödinger equation, ensembles of small things – via statistical mechanics and related fluctuation theorems, through to big things – via classical mechanics. These descriptions are complemented with a Bayesian mechanics for autonomous or active things. Although this work provides a formulation of every ‘thing’, its main contribution is to examine the implications of Markov blankets for self- organisation to nonequilibrium steady-state. In brief, we recover an information geometry and accompanying free energy principle that allows one to interpret the internal states of something as representing or making inferences about its external states. The ensuing Bayesian mechanics is compatible with quantum, statistical and classical mechanics and may offer a formal description of lifelike particles. Key words: self-organisation; nonequilibrium steady-state; active inference; active particles; free
    [Show full text]
  • The Free-Energy Principle: a Rough Guide to the Brain?
    Opinion The free-energy principle: a rough guide to the brain? Karl Friston The Wellcome Trust Centre for Neuroimaging, University College London, Queen Square, London WC1N 3BG, UK This article reviews a free-energy formulation that The motivation for the free-energy principle is again advances Helmholtz’s agenda to find principles of brain simple but fundamental. It rests upon the fact that self- function based on conservation laws and neuronal organising biological agents resist a tendency to disorder energy. It rests on advances in statistical physics, theor- and therefore minimize the entropy of their sensory states etical biology and machine learning to explain a remark- [4]. Under ergodic assumptions, this entropy is: able range of facts about brain structure and function. We Z could have just scratched the surface of what this formu- HðyÞ¼ pðyjmÞln pðyjmÞdy lation offers; for example, it is becoming clear that the ZT Bayesian brain is just one facet of the free-energy principle 1 and that perception is an inevitable consequence of active ¼ lim À ln pðyjmÞdt (Equation 1) T !1T exchange with the environment. Furthermore, one can 0 see easily how constructs like memory, attention, value, See Box 1 for an explanation of the variables and Ref. [5] reinforcement and salience might disclose their simple for details. This equation (Equation 1) means that mini- relationships within this framework. Introduction The free-energy (see Glossary) principle is a simple pos- Glossary tulate with complicated implications. It says that any [Kullback-Leibler] divergence: information divergence, information gain, cross adaptive change in the brain will minimize free-energy.
    [Show full text]
  • The Free-Energy Principle: a Unified Brain Theory?
    REVIEWS The free-energy principle: a unified brain theory? Karl Friston Abstract | A free-energy principle has been proposed recently that accounts for action, perception and learning. This Review looks at some key brain theories in the biological (for example, neural Darwinism) and physical (for example, information theory and optimal control theory) sciences from the free-energy perspective. Crucially, one key theme runs through each of these theories — optimization. Furthermore, if we look closely at what is optimized, the same quantity keeps emerging, namely value (expected reward, expected utility) or its complement, surprise (prediction error, expected cost). This is the quantity that is optimized under the free-energy principle, which suggests that several global brain theories might be unified within a free-energy framework. Free energy Despite the wealth of empirical data in neuroscience, Motivation: resisting a tendency to disorder. The An information theory measure there are relatively few global theories about how the defining characteristic of biological systems is that that bounds or limits (by being brain works. A recently proposed free-energy principle they maintain their states and form in the face of a greater than) the surprise on for adaptive systems tries to provide a unified account constantly changing environment3–6. From the point sampling some data, given a of action, perception and learning. Although this prin- of view of the brain, the environment includes both generative model. ciple has been portrayed as a unified brain theory1, its the external and the internal milieu. This maintenance Homeostasis capacity to unify different perspectives on brain function of order is seen at many levels and distinguishes bio- The process whereby an open has yet to be established.
    [Show full text]
  • A Free Energy Principle for the Brain
    Journal of Physiology - Paris 100 (2006) 70–87 www.elsevier.com/locate/jphysparis A free energy principle for the brain Karl Friston *, James Kilner, Lee Harrison The Wellcome Department of Imaging Neuroscience, Institute of Neurology, University College London, 12 Queen Square, London WC1N 3B, United Kingdom Abstract By formulating Helmholtz’s ideas about perception, in terms of modern-day theories, one arrives at a model of perceptual inference and learning that can explain a remarkable range of neurobiological facts: using constructs from statistical physics, the problems of infer- ring the causes of sensory input and learning the causal structure of their generation can be resolved using exactly the same principles. Furthermore, inference and learning can proceed in a biologically plausible fashion. The ensuing scheme rests on Empirical Bayes and hierarchical models of how sensory input is caused. The use of hierarchical models enables the brain to construct prior expectations in a dynamic and context-sensitive fashion. This scheme provides a principled way to understand many aspects of cortical organisation and responses. In this paper, we show these perceptual processes are just one aspect of emergent behaviours of systems that conform to a free energy principle. The free energy considered here measures the difference between the probability distribution of environmental quantities that act on the system and an arbitrary distribution encoded by its configuration. The system can minimise free energy by changing its con- figuration to affect the way it samples the environment or change the distribution it encodes. These changes correspond to action and perception respectively and lead to an adaptive exchange with the environment that is characteristic of biological systems.
    [Show full text]
  • A Real-World Implementation of Active Inference
    Department of Electrical Engineering Signal Processing Systems Group A Real-World Implementation of Active Inference Master Thesis Burak Erg¨ul [email protected] Supervisor: Prof.dr.ir. Bert de Vries Eindhoven, April 2020 Abstract Moving towards a world with ubiquitous automation, efficient design of intelligent au- tonomous agents that are capable of adapting to dynamic environments gains traction. In this setting, active inference emerges as a contender that inherently brings together action and perception through the minimization of a single cost function, namely free energy. Being a Bayesian inference method, active inference not only encodes uncer- tainty for perception but also for action and hence, it provides a strong expediency for building real-world intelligent autonomous agents that have to deal with uncertainties naturally found in the world. Coupled with other methods such as automated generation of inference algorithms and Forney-style factor graphs, active inference offers fast design cycles, adaptability and modularity. Furthermore, in cases where a priori specification of goal priors is prohibitively difficult, Bayesian target modelling opens up active infer- ence to more complex problems and provides a higher-level means of speeding up design cycles through learning desired future observations. In order to assess active inference's capabilities and feasibility for a real-world application, we provide a proof of concept that runs on a ground-based robot in order to navigate in a specified area to find the location chosen by the user through observing user feedback. Contents 1 INTRODUCTION . .1 1.1 Research Questions . .1 2 METHODS OF ACTIVE INFERENCE AND FREE ENERGY MINI- MIZATION .
    [Show full text]